Descriptive scope of the model

The model described in the following pages is constructed on the lines outlined in the preceding sections, with a few minor modifications; for example, the AVMs operationalised in the model are flatter, and the descriptive scope of the model is much broader, but the model contains additional relatively informal attribute specifications. There are also many possible `style options' for modelling in DATR, which will not be discussed here.

The descriptive scope of the model includes the following:

1. simplexes and inflection;
2. compound types tatpurusa, dvandva and bahuvrihi;
3. informal compositional semantic interpretation;
4. phonetic interpretation (pre- and postmorphophonemic representations);
5. orthographic interpretation (pre- and postmorphographemic representations);
6. compositionality generalised for meaning and surface interpretation at all ranks;
7. morphophonological finite state transducer;
8. morphographemic finite state transducer.

 

Term: Description:

cat category, cf. `CAT' in AVM

compound morphological category specification

graph morphographemic interpretation

head cf. `HEAD' in AVM

mass value for mass noun

modi cf. `MODI' in AVM

morph morphological attribute

operator compositionality operator, cf. `RELN' in AVM

orth orthographic (post-morphographemic) interpretation

phon phonetic interpretation (including stress marks)

plur plural inflection

mean semantic interpretation, cf. AVM `MEAN'

sing singular inflection (default value)

stem morphological category specification

stress lexical stress

surf surface interpretation (default is morphophonemic)

plain inflectional status of modifier

The morphophonological and morphographemic finite state transducers demonstrate how one formalism can be used to operationalise different theories, in this case not as an AVM based theory modelled with directed acyclic graphs, but as automata of the kind used in two-level morphology [Koskenniemi 1983], modelled with directed cyclic graphs. The terms are used are listed in Table 4.

Not all aspects of the model can be discussed in the present context, but some of the lexical specifications which can be inferred by application of the inheritance rules are illustrated here with ther synthetic compound busdriver:

Busdriver:<surf graph>           = bus-drive+er.
Busdriver:<surf graph orth>      = bus-driver.
Busdriver:<surf>                 = bVs#draiv+@.
Busdriver:<surf phon>            = //bVs/draiv@.
Busdriver:<mean>                 = {{{one_OF_{agent|instrument}}
                                     _CAN_{{action_OF_{move_vehicle}}}}
                                           _AFFECT_{{one_OF_{public_road_vehicle}}}}.
Busdriver:<plur surf graph>      = bus-drive+er#+s.
Busdriver:<plur surf graph orth> = bus-drivers.
Busdriver:<plur surf>            = bVs#draiv+@#+/Z.
Busdriver:<plur surf phon>       = //bVs/draiv@z.
Busdriver:<plur mean>            = {{{more_than_one_OF_{agent|instrument}}
                                     _CAN_{{action_OF_{move_vehicle}}}}
                                           _AFFECT_{{one_OF_{public_road_vehicle}}}.